blackhole33 / Image2text

Model discription and Inference

Image to Text modeli bu asosan pre-trained qilingan model ustiga fine-tuned qilindi juda kam dataset bilan.

epoch soni : 50 ta

loss: 0.03....

train_time: o'rtacha 45 minute.

test

Juda ham kam dataset bilan fine-tuned qilingani uchun , ko'rsatilgan dataset imagelaridan foydalanish tafsiya qilaman.

Dataset image and uning dscription holatidan bo'ladi.

misol uchun :

    from datasets import load_dataset 
    dataset = load_dataset("ybelkada/football-dataset", split="train")

Usage model

from transformers import AutoProcessor, BlipForConditionalGeneration

processor = AutoProcessor.from_pretrained("ai-nightcoder/Image2text")
model = BlipForConditionalGeneration.from_pretrained("ai-nightcoder/Image2text")

image olamiz

example = dataset[0]
image = example["image"]
image

generate qismi.

inputs = processor(images=image, return_tensors="pt").to(device)
pixel_values = inputs.pixel_values
    
generated_ids = model.generate(pixel_values=pixel_values, max_length=50)
generated_caption = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(generated_caption)

Yuqorida ko'rsatgan tartibda modeldan foydalanishni tavsiya qilaman.